OS X: Apple's Opportunity

Apple has never had a strong lobby in corporate IT departments. At most there were a few enthusiasts, if that.

A simplified history

Most of the reasons for this are PC history. Managers bought Apple IIs (with 48k RAM) with VisiCalc to do spreadsheets in the late 70s. Many IIs were not even declared as computers so that the managers could avoid involving IT departments.

This market dried up when the IBM PC arrived. Managers were still buying directly, but they started demanding connections to the corporate computers so the data could go straight to the PCs.

The IT departments saw their chance. They owned the data, the relationship with IBM (mainframes ruled), and could say how and what PCs were going to be connected to their network. They regained control except where divisions had already put together an infrastructure. Once IT departments worked with PCs, they realised just how much support was needed. This allowed IT departments to expand. Since corporate status depended on department size, this was a good thing (for them).

When Macs arrived, they didn't fit into this at all. They needed less training and less support from the IT department. Fortunately for IT departments, Apple and the dealer network knew little about selling to the corporate world. If Apple had known, Macs would now be the standard. In those days most IT departments were well known for delivering systems late, over budget, and not quite fitting user requirements.

Twenty years on, most IT departments are split between Windows people and Unix people. Nearly all the really powerful systems are Unix-based. Even in the mainframe world, where MVS still rules, 11% of IBM's last quarter sales were for Linux.

IT departments are now under pressure to cut costs. If they don't, more and more companies are prepared to outsource computing support.

Most of the current IT Directors come from the large systems world and are used to more security than is available in Windows. They can't be happy with the time spent on applying security patches to Windows servers, the time required to support PCs, and the XP pricing structure.

The time is ripe for Apple to start working directly - and through others - with IT departments.

The other Unix Vendors

Sun has been the dominant player for the past few years. With the fall off in dot-com business and, more recently, cuts in the finance sector, Sun has been bleeding red ink. This year it was forced to lay off staff for the first time. However, it is still the leading Unix vendor and has very strong relationships with the major software companies like Oracle.

The low end Sun Blade at $995 competes in price with PCs. Sun kept clear of Linux until last year when it bought Cobalt, whose popular Qube server caused Apple some naming problems. The new version of Sun's operating system, Solaris 9, will contain Linux-friendly features. Last week Sun announced it will sell x86/Linux PCs.

Sun will be using AMD chips, in a move to further establish itself in the Linux market and to slow or stop any widespread adoption of Intel's Itanium architecture. If the latter works, Sun and IBM are best placed to benefit in the server market, and it should help Apple in the workstation market.

IBM has been beating the Linux drum loudly for a couple of years and has used Linux as a way of outflanking Sun. The Power4 (PowerPC) has won major supercomputer bids and IBM has generally seen a rise in market share with the latest offerings based on this architecture. It has also hedged its chip position, in case Itanium and successors become widespread, by porting its Unix, AIX.

IBM has been doing particularly well in the server consolidation market. In this a large server or mainframe runs many virtual servers and acts as a replacement for a considerable number of low-end servers (usually Linux or Unix). As this can save considerably on support manpower, this is one of the ways IT departments can cut costs.

HP will be distracted for at least the next 12 months. Either the takeover of Compaq goes ahead, in which case integration issues will take up management time, or the takeover is voted down and a new management team will be installed. HP helped Intel with the design of the Itanium architecture and has been promoting it as the replacement for HP's own PA-RISC. HP customers, though, will probably wait until the takeover dust has settled before considering whether to adopt it.

Compaq, too, is waiting to see what happens. Its own RISC architecture, Alpha, was sold to Intel, so Compaq is hoping that McKinley (the Itanium chip replacement due 3Q02) has the performance to attract more customers. With the takeover, the limited future for Alpha and the uncertain prospects of McKinley, Compaq Unix customers must be looking at alternatives.

SGI probably has most to lose from any failure of Intel's Itanium architecture. SGI has been weak financially ever since the failed merger with Cray. Integration of the companies took management's eye off the ball at a critical time, and SGI has lost money almost ever since. While it still makes leading edge systems at the high end (for example 512 MIPS processor based SMP systems) it has been promoting Linux64/Itanium workstations. It may not now have the resources or the customer loyalty to successfully launch a new range of workstations based on a new architecture.

There has been talk about Apple buying SGI. There is no sound commercial logic behind this. SGI has too many employees for Apple to swallow comfortably. The leading edge graphics systems and SMPs are MIPS based, and MIPS has a lot less secure future than PowerPC. SGI has lost much of its Hollywood customer base, and in the last quarter it sold a large portfolio of graphics Intellectual Property Rights (IPR) to Microsoft. Also SGI's current level of losses would leave Apple struggling to make a profit.

Of the U.S. manufacturers, IBM and Sun are well placed, providing Unix continues to hold its ground against Microsoft. Outside the U.S., Fujitsu (Fujitsu-Siemens in Europe), Hitachi, and NEC are important vendors. However, with the exception of supercomputers, their major Unix computers are based on U.S. technology.

Many of the systems are sold to smaller companies by Value Added Resellers (VARs), together with their own software, software customisation etc. As the industry moves increasingly towards Linux and Solaris, VARs will no longer have such a high cost barrier to keep them selling one manufacturer's computers.

Apple's opportunity

Apple owns the best user interface in the Unix world. This gives Apple the chance to dominate a significant desktop market in corporations.

Microsoft's refusal to allow dual-boot computers means that no other OS can be installed on a PC that is sold with Windows installed. Unless users are willing to spend much of the day rebooting, many Unix users are forced to have at least two systems on their desktop - a PC for company standard programs (Microsoft Office, etc.) and a Unix workstation.

A Mac with OS X can replace these, leading to considerable savings in hardware costs and IT support costs. The user also benefits from much more space and not having to spend time transferring files between the systems.

Also playing into Apple's hands is Microsoft's refusal to port Office and other software to Linux. Until organisations are willing to accept Star Office and/or Lindows gets off the ground, Linux can not compete with OS X and Windows in the corporate desktop market.

Replacing two systems with a Mac should allow Apple to reestablish its presence in large technology companies, such as Boeing, where Macs used to be widely used. This time Apple can work with the IT department instead of against it.

Some industries are already largely Unix. Clearly Apple has had its eye on Bioinformatics for some time, with the appointment of Arthur Levinson, CEO of Genentech to the board in August 2000. Recently Genentech said it would buy 1000 iMacs and, Apple/Genentech announced a version of BLAST optimised for OS X for protein and DNA searches.

For industries like this, OS X gives companies an easy way to standardise on Unix - or at least restrict Windows to enclaves like accounting and therefore save money supporting a second OS.

Alliances with Sun and IBM are possible as Apple lacks a strong server business.

Sun has been obsessed with Microsoft for years. Scott McNeely, Sun's chairman, said, "the enemy of my enemy is also my friend." Using Microsoft products on Apple to keep out Windows may well appeal. Last week Apple, Sun, and Ericsson announced a marketing relationship for using QuickTime to stream video to mobile phones.

IBM's PC business has been losing money for years. It has dropped out of retail. It will keep selling PCs because of its Linux strategy, but an alliance with Apple would cut the PC losses. Apple could also agree to buy the PowerPC chips for these Macs from IBM.

As well as increasing the profits of IBM's Microprocessor Division, this would give Apple a strong second source for G4s. Lou Gerstner, who led IBM's comeback, is retiring soon, and this alliance would be a strong move for his successor.

While Apple is exploring these alliances, it should also start finding VARs. With the changes occurring at Compaq, HP, and SGI and the general move to Linux at the server level, many VARs will be looking at their options. Selling Macs isn't a high margin business, but selling PCs with Linux is cutthroat.

Apple investors typically look at the effects of OS X on the user base because the enthusiasm of current Mac users has kept Apple profitable. With OS X, Apple can break out of its corporate niches because it offers a cheaper alternative to Wintel/Unix users. Alliances with major Unix Vendors and VARs will give Apple the corporate sales force it has never had and can only afford to put into place to a limited extent.